NVIDIA announced the availability of RTX Spark laptops, which combine RTX graphics with NVIDIA’s AI software platform in Windows PCs. The laptops are designed for content creation, gaming and local AI workloads and will be available from ASUS, Dell, HP, Lenovo, Microsoft and MSI.
RTX Spark uses a superchip with up to 128 GB of unified memory and up to 1 petaflop of FP4 AI performance for running large language models and AI agents locally. Creators, developers and gamers can use the same Windows PC for conventional applications and local AI workloads.
RTX Spark brings NVIDIA computing to portable Windows PCs
RTX Spark uses a superchip that combines an NVIDIA Blackwell RTX GPU with an NVIDIA Grace CPU. The platform supports up to 128 GB of unified memory and up to 1 petaflop of FP4 AI performance, allowing creative applications and local AI models to sThe laptops offer up to 128 GB unified memory, 1 petaflop FP4 performance and support for local AI agents and 8K video.hare system memory.
MediaTek collaborated with NVIDIA on the CPU design, including its power, performance and connectivity characteristics.
The platform supports NVIDIA CUDA, DLSS, OptiX, Reflex, G-SYNC, TensorRT and hardware-accelerated video processing. Laptop designs are intended for portable use and include battery-powered operation.
Local AI agents support content creation and software development
RTX Spark systems can run local AI agents for tasks such as generating assets, writing code and performing other software operations.
Guided setup tools and installers allow users to install agents including GitHub Copilot, Hermes, OpenClaw, Perplexity, Pi and WorkBuddy without using a command-line interface.
NVIDIA’s Nemotron 3.5 Bolt 74B-A9B 24GB model is optimized for tool calling and multistep tasks. The 74-billion-parameter model is available on supported RTX Spark configurations.
The NVIDIA OpenShell secure runtime and Microsoft Execution Containers provide controls for agent permissions. NVIDIA agent skills allow compatible agents to interact with tools such as NVIDIA Broadcast, RTX Remix and Omniverse libraries. For example, agents can adjust Broadcast effects during a stream or process recorded clips locally.
NVIDIA PAIR allows RTX Spark systems to provide AI inference to compatible Windows, Linux and macOS computers on a home network. This can allow other computers to remain available for applications such as content creation or gaming.
Support for content creation and gaming
RTX Spark laptops support 8K 4:2:2 video editing and rendering of 3D scenes with file sizes up to 90 GB, including during battery-powered operation.
Adobe is adding NVIDIA TensorRT acceleration to AI features in Photoshop, Premiere and Substance 3D. In Premiere, the acceleration applies to functions including playback, rendering, effects and color grading. Photoshop and Premiere have native RTX Spark support through beta releases available at launch, while Substance 3D support is planned for a later release.
NVIDIA Broadcast 2.3 can remove background noise from existing recordings and modify voice output using Studio Voice Mic Profiles. Beta RTX Video features support up to 4K super resolution and 4x frame generation.
New RTX Video nodes in ComfyUI can convert 15-frame-per-second AI-generated video to 60-frame-per-second video and convert video to high dynamic range without requiring cloud inference for each use.
Local tools for coding, testing and deployment
Software development workflows often use a laptop, a cloud-based AI model and a virtual machine for testing. RTX Spark supports coding, testing and prototyping on one system.
Native CUDA support allows developers to run models and agents locally. Locally processed code and context can remain on the device, with deployment options for NVIDIA RTX AI PCs and NVIDIA GPUs in the cloud. Windows Developer Config supports setup, while Windows Subsystem for Linux provides Linux development tools.
NVIDIA Skills is an open-source GitHub repository that helps agents use NVIDIA CUDA-X tools and libraries. It is intended to support development of NVIDIA-accelerated applications without requiring extensive CUDA experience.
NVIDIA and Microsoft have prepared developer tools for RTX Spark, including a PyTorch-CUDA build for Windows on Arm, a Python build that is native by default, a native Conda installer and support for 97% of the top 15,000 Python packages. Coding agents include GitHub Copilot and Pi.
Gaming support for RTX Spark laptops
RTX Spark laptops combine portable Windows computing with NVIDIA Blackwell RTX graphics. Supported games can use NVIDIA technologies for rendering, image quality and latency management.
Launch support includes titles such as Cyberpunk 2077, Fortnite, Gears of War: E-Day, Marvel Rivals, World of Warcraft and NBA 2K27.
NBA 2K27 uses NVIDIA DLSS 5 with 3D-Guided Neural Rendering for lighting and material processing. A new AI model available at launch supports DLSS 5 frame-rate improvements of up to 1.4x. Other supported technologies include DLSS, ray tracing, G-SYNC and Reflex.
Availability
RTX Spark laptops are available globally from ASUS, Dell, HP, Lenovo, Microsoft and MSI through company websites and major retailers. Prices start at $2,599, and configurations and pricing vary by region.
The lineup includes the ASUS ProArt P14 and P16, Dell XPS 16, HP OmniBook Ultra and OmniBook X 14, Lenovo Yoga Pro 9n, Microsoft Surface Laptop Ultra, and MSI Prestige N16 Flip AI+ and Prestige N16.
For more information, visit nvidia.com.
The post NVIDIA introduces RTX Spark laptops for local AI and accelerated computing appeared first on Engineering.com.